From 5f571ca95a0cd22dcd050c43dbf111aeece89fd7 Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Bj=C3=B6rn=20Gustavsson?= Date: Mon, 26 Oct 2015 11:55:42 +0100 Subject: epp: Refactor expand_macros() As a preparation for implementing a ?FUNCTION macro, pass the entire state record to expand_macros/2 and its helpers. That will allow us to have more information available when expanding ?FUNCTION. --- lib/stdlib/src/epp.erl | 36 ++++++++++++++++++------------------ 1 file changed, 18 insertions(+), 18 deletions(-) (limited to 'lib/stdlib') diff --git a/lib/stdlib/src/epp.erl b/lib/stdlib/src/epp.erl index be7c2ec346..f55aff1e00 100644 --- a/lib/stdlib/src/epp.erl +++ b/lib/stdlib/src/epp.erl @@ -755,7 +755,7 @@ scan_toks([{'-',_Lh},{atom,_Le,elif}=Elif|Toks], From, St) -> scan_toks([{'-',_Lh},{atom,_Le,endif}=Endif|Toks], From, St) -> scan_endif(Toks, Endif, From, St); scan_toks([{'-',_Lh},{atom,_Lf,file}=FileToken|Toks0], From, St) -> - case catch expand_macros(Toks0, {St#epp.macs, St#epp.uses}) of + case catch expand_macros(Toks0, St) of Toks1 when is_list(Toks1) -> scan_file(Toks1, FileToken, From, St); {error,ErrL,What} -> @@ -763,7 +763,7 @@ scan_toks([{'-',_Lh},{atom,_Lf,file}=FileToken|Toks0], From, St) -> wait_req_scan(St) end; scan_toks(Toks0, From, St) -> - case catch expand_macros(Toks0, {St#epp.macs, St#epp.uses}) of + case catch expand_macros(Toks0, St) of Toks1 when is_list(Toks1) -> epp_reply(From, {ok,Toks1}), wait_req_scan(St#epp{macs=scan_module(Toks1, St#epp.macs)}); @@ -1145,24 +1145,24 @@ macro_expansion([T|Ts], _Anno0) -> [T|macro_expansion(Ts, T)]; macro_expansion([], Anno0) -> throw({error,loc(Anno0),premature_end}). -%% expand_macros(Tokens, Macros) +%% expand_macros(Tokens, St) %% expand_macro(Tokens, MacroToken, RestTokens) %% Expand the macros in a list of tokens, making sure that an expansion %% gets the same location as the macro call. -expand_macros(MacT, M, Toks, Ms0) -> - {Ms,U} = Ms0, +expand_macros(MacT, M, Toks, St) -> + #epp{macs=Ms,uses=U} = St, Lm = loc(MacT), Tinfo = element(2, MacT), case expand_macro1(Lm, M, Toks, Ms) of {ok,{none,Exp}} -> check_uses([{M,none}], [], U, Lm), - Toks1 = expand_macros(expand_macro(Exp, Tinfo, [], #{}), Ms0), - expand_macros(Toks1++Toks, Ms0); + Toks1 = expand_macros(expand_macro(Exp, Tinfo, [], #{}), St), + expand_macros(Toks1++Toks, St); {ok,{As,Exp}} -> check_uses([{M,length(As)}], [], U, Lm), {Bs,Toks1} = bind_args(Toks, Lm, M, As, #{}), - expand_macros(expand_macro(Exp, Tinfo, Toks1, Bs), Ms0) + expand_macros(expand_macro(Exp, Tinfo, Toks1, Bs), St) end. expand_macro1(Lm, M, Toks, Ms) -> @@ -1211,16 +1211,16 @@ get_macro_uses({M,Arity}, U) -> %% Macro expansion %% Note: io:scan_erl_form() does not return comments or white spaces. -expand_macros([{'?',_Lq},{atom,_Lm,M}=MacT|Toks], Ms) -> - expand_macros(MacT, M, Toks, Ms); +expand_macros([{'?',_Lq},{atom,_Lm,M}=MacT|Toks], St) -> + expand_macros(MacT, M, Toks, St); %% Special macros -expand_macros([{'?',_Lq},{var,Lm,'LINE'}=Tok|Toks], Ms) -> +expand_macros([{'?',_Lq},{var,Lm,'LINE'}=Tok|Toks], St) -> Line = erl_scan:line(Tok), - [{integer,Lm,Line}|expand_macros(Toks, Ms)]; -expand_macros([{'?',_Lq},{var,_Lm,M}=MacT|Toks], Ms) -> - expand_macros(MacT, M, Toks, Ms); + [{integer,Lm,Line}|expand_macros(Toks, St)]; +expand_macros([{'?',_Lq},{var,_Lm,M}=MacT|Toks], St) -> + expand_macros(MacT, M, Toks, St); %% Illegal macros -expand_macros([{'?',_Lq},Token|_Toks], _Ms) -> +expand_macros([{'?',_Lq},Token|_Toks], _St) -> T = case erl_scan:text(Token) of Text when is_list(Text) -> Text; @@ -1229,9 +1229,9 @@ expand_macros([{'?',_Lq},Token|_Toks], _Ms) -> io_lib:write(Symbol) end, throw({error,loc(Token),{call,[$?|T]}}); -expand_macros([T|Ts], Ms) -> - [T|expand_macros(Ts, Ms)]; -expand_macros([], _Ms) -> []. +expand_macros([T|Ts], St) -> + [T|expand_macros(Ts, St)]; +expand_macros([], _St) -> []. %% bind_args(Tokens, MacroLocation, MacroName, ArgumentVars, Bindings) %% Collect the arguments to a macro call. -- cgit v1.2.3 From 1604b874828e8ef992c8e17e06e7af20a0f1574b Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Bj=C3=B6rn=20Gustavsson?= Date: Mon, 26 Oct 2015 12:02:40 +0100 Subject: Implement ?FUNCTION_NAME and ?FUNCTION_ARITY macros For a long time, users have asked for one or more macros that would return the name and arity of the current function. We could define a single ?FUNCTION macro that would return a {Name,Arity} tuple. However, to access just the name or just the arity for the function, element/2 must be used. That would limit its usefulness, because element/2 is not allowed in all contexts. Therefore, it seems that we will need two macros. ?FUNCTION_NAME that expands to the name of the current function and ?FUNCTION_ARITY that expands to arity of the current function. Converting the function name to a string can be done like this: f() -> atom_to_list(?FUNCTION_NAME) ++ "/" ++ integer_to_list(?FUNCTION_ARITY). f/0 will return "f/0". The BEAM compiler will evaluate the entire expression at compile-time, so there will not be any run-time penalty for the function calls. The implementation is non-trivial because the preprocessor is run before the parser. One way to implement the macros would be to replace them with some placeholder and then let the parser or possibly a later pass replace the placeholder with correct value. That could potentially slow down the compiler and cause incompatibilities for parse transforms. Another way is to let the preprocessor do the whole job. That means that the preprocessor will have to scan the function head to find out the name and arity. The scanning of the function head can be delayed until the first occurrence of a ?FUNCTION_NAME or ?FUNCTION_ARITY. I have chosen the second way because it seems less likely to cause weird compatibility problems. --- lib/stdlib/src/epp.erl | 118 +++++++++++++++++++++++++++++++++++++++++- lib/stdlib/test/epp_SUITE.erl | 91 ++++++++++++++++++++++++++++++-- 2 files changed, 203 insertions(+), 6 deletions(-) (limited to 'lib/stdlib') diff --git a/lib/stdlib/src/epp.erl b/lib/stdlib/src/epp.erl index f55aff1e00..936c095aef 100644 --- a/lib/stdlib/src/epp.erl +++ b/lib/stdlib/src/epp.erl @@ -46,6 +46,10 @@ -type tokens() :: [erl_scan:token()]. -type used() :: {name(), argspec()}. +-type function_name_type() :: 'undefined' + | {atom(),non_neg_integer()} + | tokens(). + -define(DEFAULT_ENCODING, utf8). %% Epp state record. @@ -63,7 +67,8 @@ uses = #{} %Macro use structure :: #{name() => [{argspec(), [used()]}]}, default_encoding = ?DEFAULT_ENCODING :: source_encoding(), - pre_opened = false :: boolean() + pre_opened = false :: boolean(), + fname = [] :: function_name_type() }). %% open(Options) @@ -205,6 +210,10 @@ format_error({include,W,F}) -> io_lib:format("can't find include ~s \"~s\"", [W,F]); format_error({illegal,How,What}) -> io_lib:format("~s '-~s'", [How,What]); +format_error({illegal_function,Macro}) -> + io_lib:format("?~s can only be used within a function", [Macro]); +format_error({illegal_function_usage,Macro}) -> + io_lib:format("?~s must not begin a form", [Macro]); format_error({'NYI',What}) -> io_lib:format("not yet implemented '~s'", [What]); format_error(E) -> file:format_error(E). @@ -545,6 +554,8 @@ predef_macros(File) -> Machine = list_to_atom(erlang:system_info(machine)), Anno = line1(), Defs = [{'FILE', {none,[{string,Anno,File}]}}, + {'FUNCTION_NAME', undefined}, + {'FUNCTION_ARITY', undefined}, {'LINE', {none,[{integer,Anno,1}]}}, {'MODULE', undefined}, {'MODULE_STRING', undefined}, @@ -763,7 +774,7 @@ scan_toks([{'-',_Lh},{atom,_Lf,file}=FileToken|Toks0], From, St) -> wait_req_scan(St) end; scan_toks(Toks0, From, St) -> - case catch expand_macros(Toks0, St) of + case catch expand_macros(Toks0, St#epp{fname=Toks0}) of Toks1 when is_list(Toks1) -> epp_reply(From, {ok,Toks1}), wait_req_scan(St#epp{macs=scan_module(Toks1, St#epp.macs)}); @@ -1214,6 +1225,22 @@ get_macro_uses({M,Arity}, U) -> expand_macros([{'?',_Lq},{atom,_Lm,M}=MacT|Toks], St) -> expand_macros(MacT, M, Toks, St); %% Special macros +expand_macros([{'?',_Lq},{var,Lm,'FUNCTION_NAME'}=Token|Toks], St0) -> + St = update_fun_name(Token, St0), + case St#epp.fname of + undefined -> + [{'?',_Lq},Token]; + {Name,_} -> + [{atom,Lm,Name}] + end ++ expand_macros(Toks, St); +expand_macros([{'?',_Lq},{var,Lm,'FUNCTION_ARITY'}=Token|Toks], St0) -> + St = update_fun_name(Token, St0), + case St#epp.fname of + undefined -> + [{'?',_Lq},Token]; + {_,Arity} -> + [{integer,Lm,Arity}] + end ++ expand_macros(Toks, St); expand_macros([{'?',_Lq},{var,Lm,'LINE'}=Tok|Toks], St) -> Line = erl_scan:line(Tok), [{integer,Lm,Line}|expand_macros(Toks, St)]; @@ -1354,6 +1381,93 @@ expand_arg([A|As], Ts, _L, Rest, Bs) -> expand_arg([], Ts, L, Rest, Bs) -> expand_macro(Ts, L, Rest, Bs). +%%% +%%% Here follows support for the ?FUNCTION_NAME and ?FUNCTION_ARITY +%%% macros. Since the parser has not been run yet, we don't know the +%%% name and arity of the current function. Therefore, we will need to +%%% scan the beginning of the current form to extract the name and +%%% arity of the function. +%%% + +update_fun_name(Token, #epp{fname=Toks0}=St) when is_list(Toks0) -> + %% ?FUNCTION_NAME or ?FUNCTION_ARITY is used for the first time in + %% a function. First expand macros (except ?FUNCTION_NAME and + %% ?FUNCTION_ARITY) in the form. + + Toks1 = (catch expand_macros(Toks0, St#epp{fname=undefined})), + + %% Now extract the name and arity from the stream of tokens, and store + %% the result in the #epp{} record so we don't have to do it + %% again. + + case Toks1 of + [{atom,_,Name},{'(',_}|Toks] -> + %% This is the beginning of a function definition. + %% Scan the token stream up to the matching right + %% parenthesis and count the number of arguments. + FA = update_fun_name_1(Toks, 1, {Name,0}, St), + St#epp{fname=FA}; + [{'?',_}|_] -> + %% ?FUNCTION_NAME/?FUNCTION_ARITY used at the beginning + %% of a form. Does not make sense. + {var,_,Macro} = Token, + throw({error,loc(Token),{illegal_function_usage,Macro}}); + _ when is_list(Toks1) -> + %% Not the beginning of a function (an attribute or a + %% syntax error). + {var,_,Macro} = Token, + throw({error,loc(Token),{illegal_function,Macro}}); + _ -> + %% A macro expansion error. Return a dummy value and + %% let the caller notice and handle the error. + St#epp{fname={'_',0}} + end; +update_fun_name(_Token, St) -> + St. + +update_fun_name_1([Tok|Toks], L, FA, St) -> + case classify_token(Tok) of + comma -> + if + L =:= 1 -> + {Name,Arity} = FA, + update_fun_name_1(Toks, L, {Name,Arity+1}, St); + true -> + update_fun_name_1(Toks, L, FA, St) + end; + left -> + update_fun_name_1(Toks, L+1, FA, St); + right when L =:= 1 -> + FA; + right -> + update_fun_name_1(Toks, L-1, FA, St); + other -> + case FA of + {Name,0} -> + update_fun_name_1(Toks, L, {Name,1}, St); + {_,_} -> + update_fun_name_1(Toks, L, FA, St) + end + end; +update_fun_name_1([], _, FA, _) -> + %% Syntax error, but never mind. + FA. + +classify_token({C,_}) -> classify_token_1(C); +classify_token(_) -> other. + +classify_token_1(',') -> comma; +classify_token_1('(') -> left; +classify_token_1('{') -> left; +classify_token_1('[') -> left; +classify_token_1('<<') -> left; +classify_token_1(')') -> right; +classify_token_1('}') -> right; +classify_token_1(']') -> right; +classify_token_1('>>') -> right; +classify_token_1(_) -> other. + + %%% stringify(Ts, L) returns a list of one token: a string which when %%% tokenized would yield the token list Ts. diff --git a/lib/stdlib/test/epp_SUITE.erl b/lib/stdlib/test/epp_SUITE.erl index 955b482313..bd8939ccca 100644 --- a/lib/stdlib/test/epp_SUITE.erl +++ b/lib/stdlib/test/epp_SUITE.erl @@ -27,7 +27,7 @@ pmod/1, not_circular/1, skip_header/1, otp_6277/1, otp_7702/1, otp_8130/1, overload_mac/1, otp_8388/1, otp_8470/1, otp_8562/1, otp_8665/1, otp_8911/1, otp_10302/1, otp_10820/1, - otp_11728/1, encoding/1, extends/1]). + otp_11728/1, encoding/1, extends/1, function_macro/1]). -export([epp_parse_erl_form/2]). @@ -70,7 +70,7 @@ all() -> not_circular, skip_header, otp_6277, otp_7702, otp_8130, overload_mac, otp_8388, otp_8470, otp_8562, otp_8665, otp_8911, otp_10302, otp_10820, otp_11728, - encoding, extends]. + encoding, extends, function_macro]. groups() -> [{upcase_mac, [], [upcase_mac_1, upcase_mac_2]}, @@ -843,8 +843,9 @@ otp_8130(Config) when is_list(Config) -> "t() -> ?a.\n"), ?line {ok,Epp} = epp:open(File, []), PreDefMacs = macs(Epp), - ['BASE_MODULE','BASE_MODULE_STRING','BEAM','FILE','LINE', - 'MACHINE','MODULE','MODULE_STRING'] = PreDefMacs, + ['BASE_MODULE','BASE_MODULE_STRING','BEAM','FILE', + 'FUNCTION_ARITY','FUNCTION_NAME', + 'LINE','MACHINE','MODULE','MODULE_STRING'] = PreDefMacs, ?line {ok,[{'-',_},{atom,_,file}|_]} = epp:scan_erl_form(Epp), ?line {ok,[{'-',_},{atom,_,module}|_]} = epp:scan_erl_form(Epp), ?line {ok,[{atom,_,t}|_]} = epp:scan_erl_form(Epp), @@ -1491,6 +1492,88 @@ extends(Config) -> [] = run(Config, Ts), ok. +function_macro(Config) -> + Cs = [{f_c1, + <<"-define(FUNCTION_NAME, a).\n" + "-define(FUNCTION_ARITY, a).\n" + "-define(FS,\n" + " atom_to_list(?FUNCTION_NAME) ++ \"/\" ++\n" + " integer_to_list(?FUNCTION_ARITY)).\n" + "-attr({f,?FUNCTION_NAME}).\n" + "-attr2(?FS).\n" + "-file(?FUNCTION_ARITY, 1).\n" + "f1() ?FUNCTION_NAME/?FUNCTION_ARITY.\n" + "f2(?FUNCTION_NAME.\n">>, + {errors,[{1,epp,{redefine_predef,'FUNCTION_NAME'}}, + {2,epp,{redefine_predef,'FUNCTION_ARITY'}}, + {6,epp,{illegal_function,'FUNCTION_NAME'}}, + {7,epp,{illegal_function,'FUNCTION_NAME'}}, + {8,epp,{illegal_function,'FUNCTION_ARITY'}}, + {9,erl_parse,["syntax error before: ","f1"]}, + {10,erl_parse,["syntax error before: ","'.'"]}], + []}}, + + {f_c2, + <<"a({a) -> ?FUNCTION_NAME.\n" + "b(}{) -> ?FUNCTION_ARITY.\n" + "c(?FUNCTION_NAME, ?not_defined) -> ok.\n">>, + {errors,[{1,erl_parse,["syntax error before: ","')'"]}, + {2,erl_parse,["syntax error before: ","'}'"]}, + {3,epp,{undefined,not_defined,none}}], + []}}, + + {f_c3, + <<"?FUNCTION_NAME() -> ok.\n" + "?FUNCTION_ARITY() -> ok.\n">>, + {errors,[{1,epp,{illegal_function_usage,'FUNCTION_NAME'}}, + {2,epp,{illegal_function_usage,'FUNCTION_ARITY'}}], + []}} + ], + + [] = compile(Config, Cs), + + Ts = [{f_1, + <<"t() -> {a,0} = a(), {b,1} = b(1), {c,2} = c(1, 2),\n" + " {d,1} = d({d,1}), {foo,1} = foo(foo), ok.\n" + "a() -> {?FUNCTION_NAME,?FUNCTION_ARITY}.\n" + "b(_) -> {?FUNCTION_NAME,?FUNCTION_ARITY}.\n" + "c(_, (_)) -> {?FUNCTION_NAME,?FUNCTION_ARITY}.\n" + "d({?FUNCTION_NAME,?FUNCTION_ARITY}=F) -> F.\n" + "-define(FOO, foo).\n" + "?FOO(?FOO) -> {?FUNCTION_NAME,?FUNCTION_ARITY}.\n">>, + ok}, + + {f_2, + <<"t() ->\n" + " A = {a,[<<0:24>>,#{a=>1,b=>2}]},\n" + " 1 = a(A),\n" + " ok.\n" + "a({a,[<<_,_,_>>,#{a:=1,b:=2}]}) -> ?FUNCTION_ARITY.\n">>, + ok}, + + {f_3, + <<"-define(FS,\n" + " atom_to_list(?FUNCTION_NAME) ++ \"/\" ++\n" + " integer_to_list(?FUNCTION_ARITY)).\n" + "t() ->\n" + " {t,0} = {?FUNCTION_NAME,?FUNCTION_ARITY},\n" + " \"t/0\" = ?FS,\n" + " ok.\n">>, + ok}, + + {f_4, + <<"-define(__, _, _).\n" + "-define(FF, ?FUNCTION_NAME, ?FUNCTION_ARITY).\n" + "a(?__) -> 2 = ?FUNCTION_ARITY.\n" + "b(?FUNCTION_ARITY, ?__) -> ok.\n" + "c(?FF) -> ok.\n" + "t() -> a(1, 2), b(3, 1, 2), c(c, 2), ok.\n">>, + ok} + ], + [] = run(Config, Ts), + + ok. + check(Config, Tests) -> eval_tests(Config, fun check_test/2, Tests). -- cgit v1.2.3